Abstract: In view of the current research on mine fire escape, which mainly focuses on the smoke and heat production of the fire but ignores the impact of the roadway environment on personnel, the concept of risk degree is introduced, which is used to quantitatively describe the difficulty of the escape path of each roadway during the catastrophe period. The risk degree is divided into static risk degree and dynamic risk degree, of which the static risk degree characterizes the different slope and unevenness of the roadway, and the dynamic risk degree characterizes the CO volume fraction, roadway temperature and visibility. The FDS fire numerical simulation software was used to establish a roadway model to obtain the smoke and heat production data during the development of the fire. By obtaining the total oxygen consumption of the human body in the process of escape, the risk of personnel was characterized, and the optimal path was selected under the premise of survival conditions, providing data and theoretical guidance for emergency rescue. The results show that the route selection method of personnel escape based on the risk degree is more reasonable, which is conducive to reducing the pressure of personnel escape and reducing the casualties of blind escape.